!!! warning Filtering is a powerful tool, but it is also a double-edged sword. It is easy to make mistakes in the filter configuration. Also, since aws-nuke is in continuous development, there is always a possibility to introduce new bugs, no matter how careful we review new code.
Filtering is used to exclude or include resources from being deleted. This is important for a number of reasons to include but limited to removing the user that runs the tool.
!!! note
Filters are OR'd
together. This means that if a resource matches any filter, it will be excluded from deletion.
Currently, there is no way to do AND'ing
of filters.
Filters are traditionally done against a specific resource. However, __global__
as been introduced as a unique
resource type that can be used to apply filters to all defined resources. It's all or nothing, global cannot be used to
against some resources and not others.
Global works by taking all filters defined under __global__
and prepends to any filters found for a resource type. If
a resource does NOT have any filters defined, the __global__
ones will still be used.
Filters can only be defined under presets or account level configurations.
presets:
example:
filters:
__global__:
- property: tag:Name
value: "aws-nuke"
!!! important
Filter groups are an experimental feature and are disabled by default. To enable filter groups, use the
--feature-flag filter-groups
flag.
Filter groups are used to group filters together. This is useful when filters need to be AND'd together. For example,
if you want to delete all resources that are tagged with env:dev
and namespace:test
you can use the following filter
group:
presets:
example:
filters:
ResourceType:
- property: tag:env
value: dev
group: group1
- property: tag:namespace
value: test
group: group2
In this example, the group1
and group2
filters are AND'd together. This means that a resource must match both filters
to be excluded from deletion.
Only a single filter in a group is required to match. This means that if a resource matches any filter in a group it will count as a match for the group.
In this example, we are ignoring all resources that have the tag aws-nuke
set to ignore
. Additionally filtering
a specific instance by its id
. When the EC2Instance
resource is processed, it will have both filters applied. These
presets:
example:
filters:
__global__:
- property: tag:aws-nuke
value: "ignore"
EC2Instance:
- "i-01b489457a60298dd"
This will ultimately render as the following filters for the EC2Instance
resource:
presets:
example:
filters:
EC2Instance:
- "i-01b489457a60298dd"
- property: tag:aws-nuke
value: "ignore"
The following are comparisons that you can use to filter resources. These are used in the configuration file.
exact
contains
glob
regex
dateOlderThan
dateOlderThanNow
To use a non-default comparison type, it is required to specify an object with type
and value
instead of the
plain string.
These types can be used to simplify the configuration. For example, it is possible to protect all access keys of a
single user by using glob
:
presets:
example:
filters:
IAMUserAccessKey:
- type: glob
value: "admin -> *"
The identifier must exactly match the given string. This is the default.
Exact is just that, an exact match to a resource. The following examples are identical for the exact
filter.
presets:
example:
filters:
IAMUser:
- AWSNukeUser
- type: exact
value: AWSNukeUser
The contains
filter is a simple string contains match. The following examples are identical for the contains
filter.
presets:
example:
filters:
IAMUser:
- type: contains
value: Nuke
The identifier must match against the given glob pattern. This means the string might contain
wildcards like *
and ?
. Note that globbing is designed for file paths, so the wildcards do not match the directory
separator (/
). Details about the glob pattern can be found in the library documentation
presets:
example:
filters:
IAMUser:
- type: glob
value: "AWSNuke*"
The identifier must match against the given regular expression. Details about the syntax can be found in the library documentation.
presets:
example:
filters:
IAMUser:
- type: regex
value: "AWSNuke.*"
!!! warning You likely do not want this filter, instead you likely want dateOlderThanNow
This works by parsing the specified property into a timestamp and comparing it to the current time minus the specified
duration. The duration is specified in the value
field. The duration syntax is based on golang's duration syntax.
ParseDuration parses a duration string. A duration string is a possibly signed sequence of decimal numbers, each with optional fraction and a unit suffix, such as "300ms", "-1.5h" or "2h45m". Valid time units are "ns", "us" (or "µs"), "ms", "s", "m", "h".
Full details on duration syntax can be found in the time library documentation.
The value from the property is parsed as a timestamp and the following are the supported formats:
2006-01-02
2006/01/02
2006-01-02T15:04:05Z
2006-01-02T15:04:05.999999999Z07:00
2006-01-02T15:04:05Z07:00
In the follow example we are filtering EC2 Images that have a CreationDate
older than 1 hour.
presets:
example:
filters:
EC2Image:
- type: dateOlderThan
property: CreationDate
value: 1h
!!! note
Typically this filter is used in conjunction with invert: true
as the primary use case is to find resources
older than a date and NOT filtering them out, and instead filtering anything newer than now minus the duration
provided in the value
field of the property.
Unlike dateOlderThan
, this filter uses the property's value, assumed to be a date, compared against the current now
time modified by the duration provided in the value of the filter.
The value
in the filter must be a golang time duration value, and it is
added (if positive) or subtracted (if negative) from the current time and then the value of the property is compared
to the modified time. Note: you almost always want the value to be negative.
ParseDuration parses a duration string. A duration string is a possibly signed sequence of decimal numbers, each with optional fraction and a unit suffix, such as "300ms", "-1.5h" or "2h45m". Valid time units are "ns", "us" (or "µs"), "ms", "s", "m", "h".
presets:
example:
filters:
IAMRole:
- type: dateOlderThanNow
property: LastUsedDate
value: -12h
invert: true
If the current time is 2024-10-15T00:00:00Z
, then the modified now time is 2024-10-14T12:00:00Z
.
If the value of LastUsedDate
is 2024-10-14T14:30:00Z
then the result of the filter will be true
. It is NOT
older than the modified time, and since the invert is set to true, anything newer to the modified time is filtered.
If the value of LastUsedDate
is 2024-10-13T12:30:00Z
then the result of the filter will be false
and the resource
will be marked for removal.
By default, when writing a filter if you do not specify a property, it will use the Name
property. However, resources
that do no support Properties, aws-nuke will fall back to what is called the Legacy String
, it's essentially a
function that returns a string representation of the resource.
Some resources support filtering via properties. When a resource support these properties, they will be listed in the output like in this example:
global - IAMUserPolicyAttachment - 'admin -> AdministratorAccess' - [RoleName: "admin", PolicyArn: "arn:aws:iam::aws:policy/AdministratorAccess", PolicyName: "AdministratorAccess"] - would remove
To use properties, it is required to specify an object with properties
and value
instead of the plain string.
These types can be used to simplify the configuration. For example, it is possible to protect all access keys of a single user:
presets:
example:
filters:
IAMUserAccessKey:
- property: UserName
value: "admin"
Any filter result can be inverted by using invert: true
, for example:
presets:
example:
filters:
CloudFormationStack:
- property: Name
value: "foo"
invert: true
In this case any CloudFormationStack but the ones called "foo" will be filtered. Be aware that aws-nuke internally takes every resource and applies every filter on it. If a filter matches, it marks the node as filtered.
It is also possible to use Filter Properties and Filter Types together. For example to protect all Hosted Zone of a specific TLD:
presets:
example:
filters:
Route53HostedZone:
- property: Name
type: glob
value: "*.rebuy.cloud."
It is possible to filter this is important for not deleting the current user for example or for resources like S3
Buckets which have a globally shared namespace and might be hard to recreate. Currently, the filtering is based on
the resource identifier. The identifier will be printed as the first step of aws-nuke (eg i-01b489457a60298dd
for an EC2 instance).
!!! warning Even with filters you should not run aws-nuke on any AWS account, where you cannot afford to lose all resources. It is easy to make mistakes in the filter configuration. Also, since aws-nuke is in continuous development, there is always a possibility to introduce new bugs, no matter how careful we review new code.
The filters are part of the account-specific configuration and are grouped by resource types. This is an example of a
config that deletes all resources but the admin
user with its access permissions and two access keys:
---
regions:
- global
- us-east-1
account-blocklist:
- 1234567890
accounts:
0987654321:
filters:
IAMUser:
- "admin"
IAMUserPolicyAttachment:
- "admin -> AdministratorAccess"
IAMUserAccessKey:
- "admin -> AKSDAFRETERSDF"
- "admin -> AFGDSGRTEWSFEY"
Any resource whose resource identifier exactly matches any of the filters in the list will be skipped. These will be marked as "filtered by config" on the aws-nuke run.
It might be the case that some filters are the same across multiple accounts. This especially could happen, if provisioning tools like Terraform are used or if IAM resources follow the same pattern.
For this case aws-nuke supports presets of filters, that can applied on multiple accounts. A configuration could look like this:
---
regions:
- "global"
- "eu-west-1"
account-blocklist:
- 1234567890
accounts:
555421337:
presets:
- "common"
555133742:
presets:
- "common"
- "terraform"
555134237:
presets:
- "common"
- "terraform"
filters:
EC2KeyPair:
- "notebook"
presets:
terraform:
filters:
S3Bucket:
- type: glob
value: "my-statebucket-*"
DynamoDBTable:
- "terraform-lock"
common:
filters:
IAMRole:
- "OrganizationAccountAccessRole"
aws-nuke deletes a lot of resources and there might be added more at any release. Eventually, every resource should get deleted. You might want to restrict which resources to delete. There are multiple ways to configure this.
One way are filters, which already got mentioned. This requires to know the identifier of each resource. It is also
possible to prevent whole resource types (eg S3Bucket
) from getting deleted with two methods.
It is also possible to configure the resource types in the config file like in these examples:
regions:
- "us-east-1"
account-blocklist:
- 1234567890
resource-types:
# Specifying this in the configuration will ensure that only these three
# resources are targeted by aws-nuke during it's run.
includes:
- S3Object
- S3Bucket
- IAMRole
accounts:
555133742: {}
regions:
- "us-east-1"
account-blocklist:
- 1234567890
resource-types:
# Specifying this in the configuration will ensure that these resources
# will be specifically excluded from aws-nuke during it's run.
excludes:
- IAMUser
accounts:
555133742: {}
If includes
are specified in multiple places (e.g. CLI and account specific), then a resource type must be specified
in all places. In other words each configuration limits the previous ones.
If an exclude is used, then all its resource types will not be deleted.
Hint: You can see all available resource types with this command:
aws-nuke resource-types
It is also possible to include and exclude resources using the command line arguments:
- The
--include
flag limits nuking to the specified resource types. - The
--exclude
flag prevent nuking of the specified resource types.